Fsonicsmith, thanks. I essentially agree with everything in your post above. There are always myriad tradeoffs in a design, and the net result of those tradeoffs often will not conform to commonly stated paradigms.
For example a major reason, and perhaps the most major reason, for ARCs 20K minimum load recommendation is their use of a coupling capacitor at the outputs of their line stages and preamps, which of course is very commonly done in tube-based line-level components, and which among other things causes a substantial rise in output impedance at deep bass frequencies. And while that effect could be minimized by simply increasing the value of the capacitor (i.e., the number of microfarads), doing so would most likely increase the sonic colorations introduced by the capacitor itself, and/or result in the capacitor being a larger and potentially impractical physical size.
And an alternative approach that is used in some tube designs, using a transformer instead of a capacitor, certainly can have tradeoffs of its own.
Regarding feedback, yes, ARC power amp designs tend to use more of it than many and probably most high quality tube amps from other manufacturers. Consequently they have lower effective output impedances and higher damping factors than those other amps, which in turn can be either an advantage or a disadvantage or neither depending on the particular speaker that is being used. And the downsides feedback can potentially have with respect to the intrinsic sonic character of the amp will depend on numerous other aspects of its design.
As you aptly summarized, "it all depends."
Stringreen, I have no particular knowledge regarding the sensitivity of your Ayre components to cable differences. An experiment that could possibly be informative, though, would be to **temporarily** put a cheater plug on the amp’s power cord, to defeat its safety ground connection. If that results in a perceptible sonic difference it would suggest that ground loop effects may be present, which in turn can result in sensitivity to cable differences especially when the length involved is long. Note Ralph’s criterion no. 2 for minimizing the sensitivity of balanced interfaces to cable differences, in his post that I referenced earlier:
Atmasphere 3-22-2013
... 2) Ground is ignored- the signal occurs only between pin 2 and 3 (this is where most high end audio preamps have a problem- as soon as there are signal currents in the shield of the cable, the construction of the cable becomes critical).
Regards,
-- Al
